United States General Accounting Office
GAO
Testimony
Before the Subcommittee on Government Efficiency,
Financial Management, and Intergovernmental Relations,
Committee on Government Reform, House of
Representatives
For Release
on Delivery
Expected at 10 a.m., PDT
Wednesday
August 29, 2001
INFORMATION
SECURITY
Code Red, Code Red II, and
SirCam Attacks Highlight
Need for Proactive Measures
Statement of Keith A. Rhodes
Chief Technologist
GAO-01-1073T
Mr. Chairman and Members of the Subcommittee:
Thank you for inviting me to participate in today’s hearing on the most recent
rash of computer attacks. This is the third time I’ve testified before Congress
over the past several years on specific viruses—first, the “Melissa” virus in April
1999 and second, the “ILOVEYOU” virus in May 2000. At both hearings, I
stressed that the next attack would likely propagate faster, do more damage, and
be more difficult to detect and counter.
Again, we are having to deal with destructive attacks that are reportedly costing
billions. In the past 2 months, organizations and individuals have had to contend
with several particularly vexing attacks. The most notable, of course, is Code
Red but potentially more damaging are Code Red II and SirCam. Together, these
attacks have infected millions of computer users, shut down Web sites, slowed
Internet service, and disrupted business and government operations. They have
already caused billions of dollars of damage and their full effects have yet to be
completely assessed.
Today, I would like to discuss the makeup and potential threat that each of these
viruses pose as well as reported damages. I would also like to talk about progress
being made to protect federal operations and assets from these types of attacks
and the substantial challenges still ahead.
The Attacks
Despite some similarities, each of the recent attacks is very different in its
makeup, method of attack, and potential damage. Generally, Code Red and Code
Red II are both “worms,” which are attacks that propagate themselves through
networks without any user intervention or interaction. They both take advantage
of a flaw in a component of versions 4.0 and 5.0 of Microsoft’s Internet
Information Services (IIS) Web server software.
Code Red originally sought to do damage by defacing Web pages and by denying
access to a specific Web site by sending it massive amounts of data, which
essentially would shut it down. This is known as a denial-of-service (DoS) attack.
Code Red II is much more discreet and potentially more damaging. Other than
sharing the name of the original worm, the only similarity Code Red II has with
Code Red is that it exploits the same IIS vulnerability to propagate itself. Code
Red II installs “backdoors” on infected Web servers, making them vulnerable to
hijacking by any attacker who knows how to exploit the backdoor. It also spreads
faster than Code Red. Both attacks have the potential to decrease the speed of the
Internet and cause service disruptions. More importantly, these worms broadcast
to the Internet the servers that are vulnerable to this flaw, which allows others to
attack the servers and perform other actions that are not related to Code Red.
Page 1
GAO-01-1073T
SirCam is a malicious computer virus that spreads primarily through E-mail.
Once activated on an infected computer, the virus searches through a select
folder and mails user files acting as a “Trojan horse” to E-mail addresses in the
user’s address book. A Trojan horse, or Trojan, is a program containing hidden
code allowing the unauthorized collection, falsification, or destruction of
information. If the user’s files are sensitive in nature, then SirCam not only
succeeds in compromising the user’s computer, but also succeeds in breaching
the data’s confidentiality. In addition to spreading, the virus can attempt to delete
a victim’s hard drive or fill the remaining free space on the hard drive making it
impossible to perform common tasks such as saving files or printing. This form
of attack is extremely serious since it is one from which it is very difficult to
recover.
SirCam is much more stealthy than the Melissa and ILOVEYOU viruses
because it does not need to use the victim’s E-mail program to replicate. It has its
own internal capabilities to mail itself to other computers. SirCam also can
spread through another method. It can copy itself to other unsuspecting
computers connected through a Windows network (commonly referred to as
Windows network computers) that has been granted read/write access to the
infected computer. Like Code Red and Code Red II, SirCam can slow the
Internet. However, SirCam poses a greater threat to the home PC user than that of
the Code Red worms.
Table 1 provides a high-level comparison of the attacks. The attachment to this
testimony answers the questions in the table in greater detail.
Table 1: High-level Comparison of the Attacks
Code Red
What is it?
Code Red is a worm, which
is a computer attack that
propagates through
networks without user
intervention. This particular
How does it spread?
The worm scans the
Internet, identifies
vulnerable systems, and
infects these systems by
installing itself. Each newly
Page 2
Who is at risk?
Users with Microsoft IIS
server installed with
Windows NT version 4.0 or
Windows 2000.
What damage can it do?
The program can deface
Web sites, and was
designed to perform a DoS
attack against the
www.whitehouse.gov Web
GAO-01-1073T
Code Red II
SirCam
What is it?
worm makes use of a
vulnerability in Microsoft’s
Internet Information
Services (IIS) Web server
software—specifically, a
buffer overflow.
Code Red II is also a worm
that exploits the same IIS
vulnerability. However, the
worm also opens a
backdoor on an infected
server that allows any
follow-on remote attacker to
execute arbitrary
commands.
SirCam is a malicious
computer virus that spreads
through E-mail and
potentially through
unprotected network
connections. Once the
malicious code has been
executed on a system, it
may reveal or delete
sensitive information.
How does it spread?
installed worm joins all the
others, causing the rate of
scanning to grow rapidly.
Who is at risk?
Code Red II spreads like
Code Red; however, in
doing so, it selects Internet
addresses that are in the
same network range as the
infected computer to
increase the likelihood of
finding susceptible victims.
Users with Microsoft IIS
Web server software
installed with Windows
2000.
What damage can it do?
site. It can also decrease
the speed of the Internet.
Like Code Red, Code Red II
can decrease the speed of
the Internet. Unlike Code
Red, it also leaves the
infected system open to any
attacker who can alter or
destroy files and create a
denial of service. It does not
deface Web pages.
This mass-mailing virus
Any E-mail user or user of a SirCam can publicly release
attempts to send itself to E- computer with unprotected
sensitive information and
mail addresses found in the Windows network
delete files and folders. It
Windows Address Book and connections to the infected can also fill the remaining
addresses found in cached computer.
free space on the
browser files. It also
computer’s hard drive.
attempts to copy itself to
Furthermore, it can lead to
specific Windows
a decrease in the speed of
networked computers.
the Internet.
Systems infected by Code Red and SirCam can be fixed relatively easily. A patch
made available by Microsoft can remove the vulnerability exploited by Code Red
and rebooting the infected computer removes the worm itself. Updating and
using antivirus software can help detect and partially recover from SirCam.
Patching and rebooting an infected server is not enough when a system is hit by
Code Red II. Instead, the system’s hard drive should be reformatted, and all
software should be reinstalled to ensure that the system is free of other backdoor
vulnerabilities.
Of course, there are a number of other immediate actions organizations can take
to ward off attacks. These include:
using strong passwords,
verifying software security settings,
installing firewalls,
backing up files early and often,
ensuring that known software vulnerabilities are reduced by promptly
implementing software patches available from vendors,
ensuring that policies and controls already implemented are operating as
intended,
Page 3
GAO-01-1073T
using scanners that automatically search for system vulnerabilities,
using password-cracking tools to assess the password strength of the audited
users,
using network monitoring tools to identify suspicious network activity, and
developing and distributing lists of the most common types of vulnerabilities and
suggested corrective actions.
Impact of the Attacks
Reports from various media and computer security experts indicate that the
impact of these viruses has been extensive. On July 19, the Code Red worm
infected more than 250,000 systems in just 9 hours, according to the National
Infrastructure Protection Center (NIPC). An estimated 975,000 servers have been
infected in total, according to Computer Economics, Inc. Code Red and Code
Red II have also reportedly disrupted both government and business operations,
principally by slowing Internet service and forcing some organizations to
disconnect themselves from the Internet.
For example, reports have noted that (1) the White House had to change the
numerical Internet address that identifies its Web site to the public, (2) the
Department of Defense was forced to briefly shut down its public Web sites and
(3) Treasury’s Financial Management Service was infected and also had to
disconnect itself from the Internet. Code Red worms also reportedly hit
Microsoft’s popular free E-mail service, Hotmail; caused outages for users of
Qwest’s high-speed Internet service nationwide; and caused delays in package
deliveries by infecting systems belonging to FedEx Corp. There are also
numerous reports of infections in other countries.
The economic costs resulting from Code Red attacks are already estimated to be
over $2.4 billion.1 These involve costs associated with cleaning infected systems
and returning them to normal service, inspecting servers to determine the need
for software patches, patching and testing services as well as the negative impact
on the productivity of system users and technical staff.
Although Code Red’s reported costs have not yet surpassed damages estimated
for last year’s ILOVEYOU virus, which is now estimated to be more than $8
billion2, the Code Red attacks are reportedly more costly than 1988’s Morris
worm. This particular worm exploited a flaw in the Unix operating system and
affected VAX computers from Digital Equipment Corp. and Sun 3 computers
1
Estimate was developed by Computer Economics Inc.
2
Computer Economics, Inc.
Page 4
GAO-01-1073T
from Sun Microsystems, Inc. It was intended to only infect each computer once,
but a bug allowed it to replicate hundreds of times, crashing computers in the
process. Approximately 10 percent of the U.S. computers connected to the
Internet effectively stopped at the same time. At that time, the network had
grown to more than 88,000 computers and was a primary means of
communication among computer security experts.3
SirCam has also reportedly caused some havoc. It is allegedly responsible for the
leaking of secret documents from the government of Ukraine. And it reportedly
infected a computer at the Federal Bureau of Investigation (FBI) late last month
and sent some private, but not sensitive or classified, documents out in an E-mail.
There are reports that SirCam has surfaced in more than 100 countries.
Attacks Underscore
Challenges Involved in
Protecting Systems
GAO has identified information security as a governmentwide high risk issue
since 1997. As these incidents continue, the federal government continues to
face formidable challenges in protecting its information systems assets and
sensitive data. These include not only an ever changing and growing
sophistication in the nature of attacks but also an urgent need to strengthen
agency security controls as well as a need for a more concerted and effective
governmentwide coordination, guidance, and oversight. Today, I would like to
briefly discuss these challenges. I would also like to discuss progress that has
been made in addressing them, including improvements in agency controls,
actions to strengthen warning and crisis management capabilities, and new
legislation to provide a comprehensive framework for establishing and ensuring
effectiveness of information security controls over information resources that
support federal government operations and assets. These are positive steps
toward taking a proactive stand in protecting sensitive data and assets.
First, these latest incidents again show that computer attack tools and techniques
are becoming increasingly sophisticated. The Code Red attack was more
sophisticated than those experienced in the past because the attack combined a
worm with a denial-of-service attack. Further, with some reprogramming, each
variant of Code Red got smarter in terms of identifying vulnerable systems. Code
Red II exploited the same vulnerability to spread itself as the original Code Red.
However instead of launching a DoS attack against a specific victim, it gives an
attacker complete control over the infected system, thereby letting the attacker
perform any number of undesirable actions. SirCam was a more sophisticated
3
http://www.cert.org/encyc_article/tocencyc.html.
Page 5
GAO-01-1073T
version of the ILOVEYOU virus, no longer needing the victim’s E-mail program
to spread.
In the long run, it is likely that hackers will find ways to attack more critical
components of the Internet, such as routers and network equipment, rather than
just Web site servers or individual computers. Further, it is likely that viruses will
continue to spread faster as a result of the increasing connectivity of today’s
networks and the growing use of commercial-off-the-shelf (COTS) products,
which, once a vulnerability is discovered, can be easily exploited for attack by all
their users because of the widespread use of the products.
Second, the recent attacks foreshadow much more devastating Internet threats to
come. According to official estimates, over 100 countries already have or are
developing computer attack capabilities. Further, the National Security Agency
has determined that potential adversaries are developing a body of knowledge
about U.S. systems and methods to attack them. Meanwhile, our government and
our nation have become increasingly reliant on interconnected computer systems
to support critical operations and infrastructures, including telecommunications,
finance, power distribution, emergency services, law enforcement, national
defense, and other government services. As a result, there is a growing risk that
terrorists or hostile foreign states could severely damage or disrupt national
defense or vital public operations through computer-based attacks on the nation’s
critical infrastructures.
Third, agencies do not have an effective information security program to prevent
and respond to attacks—both external attacks, like Code Red, Code Red II, and
SirCam, and internal attempts to manipulate or damage systems and data. More
specifically, we continue to find that poor security planning and management are
the rule rather than the exception. Most agencies do not develop security plans
for major systems based on risk, have not formally documented security policies,
and have not implemented programs for testing and evaluating the effectiveness
of the controls they rely on.
Agencies also often lack effective access controls to their computer resources and
consequently cannot protect these assets against unauthorized modification, loss,
and disclosure. Moreover, application software development and change controls
are weak; policies and procedures governing segregation of duties are ineffective;
and access to the powerful programs and sensitive files associated with a
computer systems operation is not well-protected. In fact, over the past several
years, our analyses as well as those of the Inspectors General have found that
virtually all of the largest federal agencies have significant computer security
weaknesses that place critical federal operations and assets at risk to computerbased attacks.
Page 6
GAO-01-1073T
In recognition of these serious security weaknesses, we and the Inspectors
General have made recommendations to agencies regarding specific steps they
should take to make their security programs effective.4 Also, in 2001, we again
reported information security as a high-risk area across government, as we did in
our 1997 and 1999 high-risk series.5
Fourth, the government still lacks robust analysis, warning, and response
capabilities. Often, for instance, reporting on incidents has been ineffective—
with information coming too late for agencies to take proactive measures to
mitigate damage. This was especially evident in the Melissa and ILOVEYOU
attacks. There is also a lack of strategic analysis to determine the potential
broader implications of individual incidents. Such analysis looks beyond one
specific incident to consider a broader set of incidents or implications that may
indicate a potential threat of national importance.
Further, as we recently reported,6 the ability to issue prompt warnings about
attacks is impeded because of (1) a lack of a comprehensive governmentwide or
nationwide framework for promptly obtaining and analyzing information on
imminent attacks, (2) a shortage of skilled staff, (3) the need to ensure that undue
alarm is not raised for insignificant incidents, and (4) the need to ensure that
sensitive information is protected, especially when such information pertains to
law enforcement investigations underway. Lastly, government entities have not
developed fully productive information-sharing and cooperative relationships.
We recently made a variety of recommendations to the Assistant to the President
for National Security Affairs and the Attorney General regarding the need to
more fully define the role and responsibilities of the NIPC, develop plans for
establishing analysis and warning capabilities, and formalize information-sharing
relationships with the private sector and federal entities.7
4
See, for example, Information Security: Serious and Widespread Weaknesses Persist at Federal
Agencies (GAO/AIMD-00-295, September 6, 2000).
5
High-Risk Series: An Update (GAO-01-263, January 2001).
6
Critical Infrastructure Protection: Significant Challenges in Developing Analysis, Warning, and
Response Capabilities (GAO-01-1005T, July 25, 2001).
7
The NIPC agreed with generally agreed with our findings and stated that the NIPC considers it of
the utmost urgency to address the shortcomings we identified. However, the NIPC did not
comment on several key recommendations, including the need to improve cooperative relationships
with other federal entities, such as Defense and the Secret Service. See Critical Infrastructure
Protection: Significant Challenges in Developing National Capabilities (GAO-01-323, April 25,
2001).
Page 7
GAO-01-1073T
Fifth, most of the nation’s critical infrastructure is owned by the private sector.
Solutions, therefore, need to be developed and implemented in concert with the
private sector, and they must be tailored sector by sector, through consultation
about vulnerabilities, threats, and possible response strategies. Putting together
effective partnerships with the private sector is difficult, however. Disparate
interests between the private sector and the government can lead to profoundly
different views and perceptions about threats, vulnerabilities, and risks, and they
can affect the level of risk each party is willing to accept and the costs each is
willing to bear. Moreover, industry has raised concerns that it could potential
face antitrust violations for sharing information. Lastly, there is a concern that an
inadvertent release of confidential business material, such as trade secrets or
proprietary information, could damage reputations, lower consumer confidence,
hurt competitiveness, and decrease market shares of firms.
Fortunately, we are beginning to see improvements that should help agencies
ward off attacks. We reported earlier this year8 that several agencies have taken
significant steps to redesign and strengthen their information security programs.
For example, the Internal Revenue Service (IRS) has made notable progress in
improving computer security at its facilities, corrected a significant number of
identified weaknesses, and established a service-wide computer security
management program. Similarly, the Environmental Protection Agency has
moved aggressively to reduce the exposure of its systems and data and to correct
weaknesses we identified in February 2000.
Moreover, the Federal Computer Incident Response Center (FedCIRC) and the
NIPC have both expanded their efforts to issue warnings of potential computer
intrusions and to assist in responding to computer security incidents. In
responding to the Code Red and Code Red II attacks, FedCIRC and NIPC
worked together with Carnegie Mellon’s CERT Coordination Center, the Internet
Security Alliance, the National Coordinating Center for Telecommunications, the
Systems Administrators and Network Security (SANS) Institute, and other
private companies and security organizations to warn the public and encourage
system administrators and home users to voluntarily update their software.
We also recently reported on a number of other positive actions taken by NIPC to
develop analysis, warning, and response capabilities. For example, since its
establishment, the NIPC has issued a variety of analytical products to support
computer security investigations. It has established a Watch and Warning Unit
that monitors the Internet and other media 24 hours a day to identify reports of
8
High-Risk Series: An Update (GAO-01-263, January 2001).
Page 8
GAO-01-1073T
computer-based attacks. It has developed crisis management capabilities to
support a multi-agency response to the most serious incidents from FBI’s
Washington, D.C., Strategic Information Operations Center.
The administration is currently reviewing the federal strategy for critical
infrastructure protection that was originally outlined in Presidential Decision
Directive (PDD) 63, including provisions related to developing analytical and
warning capabilities that are currently assigned to the NIPC. On May 9, 2001, the
White House issued a statement saying that it was working with federal agencies
and private industry to prepare a new version of the “national plan for cyberspace
security and critical infrastructure protection” and reviewing how the government
is organized to deal with information security issues.
Lastly, the Congress recently enacted legislation to provide a comprehensive
framework for establishing and ensuring the effectiveness of information security
controls over information resources that support federal government operations
and assets. This legislation9—known as Government Information Security
Reform (GISR)—requires agencies to implement an agencywide information
security program that is founded on a continuing risk management cycle. GISR
also added an important new requirement by calling for an independent
evaluation of the information security program and practices of an agency.
These evaluations are to be used by OMB as the primary basis for its summary
report to the Congress on governmentwide information security.
In conclusion, the attacks we are dealing with now are smarter and more
threatening than the ones we were dealing with last year and the year before. But
I believe we are still just witnessing warning shots of potentially much more
damaging and devastating attacks on the nation’s critical infrastructures. To that
end, it’s vital that federal agencies and the government as a whole become
proactive rather than reactive in their efforts to protect sensitive data and assets.
In particular, as we have recommended in many reports and testimonies,10
9
Floyd D. Spence, National Defense Authorization Act for Fiscal Year 2001, P.L. 106-398, Title
X, Subtitle G, 114 Stat. 1654, 1654A-265 (2000).
10
See, for example, Information Security: Serious and Widespread Weaknesses Persist at Federal
Agencies (GAO/AIMD-00-295, September 6, 2000); Critical Infrastructure Protection:
Comprehensive Strategy Can Draw on Year 2000 Experiences (GAO/AIMD-00-1, October 1,
1999); Critical Infrastructure Protection: Comments on the National Plan for Information Systems
Protection (GAO/T-AIMD-00-72, February 1, 2000) and Critical Infrastructure Protection:
Challenges to Building A Comprehensive Strategy for Information Sharing and Coordination
(GAO/T-AIMD-00-268, July 26, 2000).
Page 9
GAO-01-1073T
agencies need more robust security planning, training, and oversight. The
government as a whole needs to fully develop the capability to strategically
analyze cyber threats and warn agencies in time for them to avert damage. It also
needs to continue building on private-public partnerships—not just to detect and
warn about attacks—but to prevent them in the first place. Most of all, trust needs
to be established among a broad range of stakeholders, roles and responsibilities
need to be clarified, and technical expertise needs to be developed. Lastly,
becoming truly proactive will require stronger leadership by the federal
government to develop a comprehensive strategy for critical infrastructure
protection, work through concerns and barriers to sharing information, and
institute the basic management framework needed to make the federal
government a model of critical infrastructure protection.
Mr. Chairman and Members of the Subcommittee, this concludes my statement.
I would be pleased to answer any questions that you or Members of the
Subcommittee may have.
Contacts and Acknowledgment
For further information, please contact Keith Rhodes at (202) 512-6412.
Individuals making key contributions to this testimony included Cristina
Chaplain, Edward Alexander, Jr., Tracy Pierson, Penny Pickett, and Chris
Martin.
Page 10
GAO-01-1073T
Enclosure I: Details on the Attacks
Code Red
Question
What is it?
Answer
Code Red is a worm, which is a computer attack that propagates through networks without
user intervention. This particular worm makes use of a vulnerability in Microsoft’s Internet
Information Services (IIS) Web server software—specifically, a buffer overflow.a The worm
looks for systems running IIS (versions 4.0 and 5.0) that have not patched the unchecked
vulnerability, and exploits the vulnerability to infect those systems.
Code Red was initially written to deface the infected computer’s Web site and to perform a
distributed denial of service (DDoS) attack against the numerical Internet address used by
www.whitehouse.gov. Two subsequent versions of Code Red do not deface Web pages but
still launch the DDoS attack.
How does it spread?
Code Red was first reported on July 17, 2001. The worm is believed to have started at a
university in Guangdong, China.
The worm scans the Internet, identifies vulnerable systems, and infects these systems by
installing itself. Each newly installed worm joins all the others causing the rate of scanning to
grow rapidly.
Who is at risk?
What damage can it do?
The first version of Code Red created a randomly generated list of Internet addresses to infect.
However, the algorithm used to generate the list was flawed, and infected systems ended up
reinfecting each other. The subsequent versions target victims a bit differently, increasing the
rate of infection.
Users with a Microsoft IIS server installed with Windows NT version 4.0 and Windows 2000.
The original variant of Code Red (CRv1) can deface the infected computer’s Web site and
used the infected computer to perform a DDoS attack against the Internet address of the
www.whitehouse.gov Web site. Subsequent variants of Code Red (CRv2a and CRv2b) no
longer defaced the infected computer’s Web site making detection of the worm harder. These
subsequent variants continued to target the www.whitehouse.gov Web site and used smarter
methods to target new computers for infection.
The uncontrolled growth in scanning can also decrease the speed of the Internet and cause
sporadic but widespread outages among all types of systems.
What can you do if you’re infected?
Specifically,

Although the initial version, CRv1, defaces the Web site, the primary impact to the server
is performance degradation as a result of the scanning activity of this worm. This
degradation can become quite severe since it is possible for a worm to infect the same
machine multiple times.

Other entities, even those that are not vulnerable to Code Red, are impacted because
servers infected by Code Red scan their systems and networks. Depending on the number
of servers performing this scan, these entities may experience network denial of service.
This was especially true with the implementation of CRv1 since a “flaw” in the random
number generator essentially targeted the same servers. As noted above, this behavior is
not found in the later variants. However, the end result may be the same since CRv2a and
CRv2b use improved randomization techniques that facilitate more prolific scanning.
Install a patch made available by Microsoft and reboot the system. (The patch should also be
installed as a preventative measure).
Page 11
GAO-01-1073T
Technical Details on How the Code Red Worm Operates
The Code Red worm has three phases – discovery and propagation, attack, and dormancy. Execution of these phases is based upon
the day of the month.
Phase 1: Discovery and
Between day 1 and day 19 of any month, Code Red performs its discovery and propagation
Propagation
function. It does this by generating 100 subprograms on an infected server. All but one of these
subprograms has the task of identifying and infecting other vulnerable Web servers by
scanning a generated list of Internet addresses. Once a target system is identified, Code Red
uses standard Web server communication to exploit the flaw and send itself to the vulnerable
server. Once a new server is infected, the process continues.
CRv1 created a randomly generated list of Internet addresses to infect. However, the algorithm
used to generate the random number list was “flawed”, and infected systems ended up reinfecting each other because the random list that each computer generated was the same.
CRv2a and CRv2b were modified to generate actual random lists of Internet addresses that
were more effective at identifying potential servers that had not already been attacked.
Therefore, these versions can ultimately infect greater numbers of unprotected servers.
CRv1 also defaced the target system’s Web site. This was done by replacing site’s actual Web
page with the message, “HELLO! Welcome to http://www.worm.com! Hacked by Chinese!”b
This message enabled system administrators to easily identify when their servers had been
infected. CRv2a and CRv2b modified the functionality so it would no longer deface Web pages,
forcing system administrators to be proactive in determining infection. Descriptions of the
variants are listed below.
Phase 2: Attack
Phase 3: Dormancy
CRv1: Web site defacement and “random” target selection for additional attacks.
CRv2a: No Web defacement and modified random target selection
CRv2b: No Web defacement and better target selection by optimizing the random number
generation process, i.e., better target addresses are generated. Due to the target
optimization, systems infected with version 2b are able to infect new systems at a faster rate
than version 2a.
Between day 20 and day 27 of any month is Code Red’s attack phase. Once Code Red
determines the date to be within this designated attack date range, each infected server
participates in a DDoS attack by sending massive amounts of data to its intended target, the
numeric Internet address of the White House Web site. Since all infected servers are set to
attack the same target on the same set of dates, the large amount of Internet traffic is expected
to flood the Internet with data and bombard a numeric address used by www.whitehouse.gov
with more data than it can handle. This flooding of data would cause the Web server to stop
responding to all Web server requests, including legitimate users surfing the White House Web
site.
From day 28 to the end of the month, the Code Red worm lays dormant, going into an infinite
sleep phase. Although the worm remains in the computer’s memory until the system is rebooted,
Code Red will not propagate or initiate any attacks once it enters dormancy. According to testing
performed by Internet Security Systems, Carnegie Mellon’s CERT Coordination Center
(CERT/CC), and the Federal Bureau of Investigation’s (FBI) National Infrastructure Protection
Center (NIPC), the dormant worm cannot be awakened to restart the process.
Code Red II
Question
What is it?
Answer
Code Red II is also a worm that makes use of a buffer overflow vulnerability in Microsoft’s IIS
Web server software.
Page 12
GAO-01-1073T
Question
Answer
Except for using the buffer overflow injection mechanism, the worm is very different than the
original Code Red and its variants. In fact, it is more dangerous because it opens backdoors on
infected servers that allow any follow-on remote attackers to execute arbitrary commands.
There is no DDoS attack function in Code Red II and, to date, no variants of Code Red II have
been discovered.
How does it spread?
Code Red II was reported on August 4 by industry analysts and government agencies.
Like Code Red, the worm scans the Internet, identifies vulnerable systems, and infects these
systems by installing itself. Each newly installed worm joins all the others causing the rate of
scanning to grow rapidly.
Who is at risk?
What damage can it do?
Code Red II, however, selects Internet addresses that look like they may be in the same network
as the infected computer to increase the likelihood of finding susceptible victims.
Users with Microsoft IIS Web server software installed with Windows 2000.
Like Code Red, Code Red II can decrease the speed of the Internet and cause sporadic but
widespread outages among all types of systems. Unlike Code Red, it also leaves the infected
system open to any attacker who can alter or destroy files and create a denial of service attack.
Specifically,
Because of the worm’s preference to target its closest neighbors, combined with the enormous
amount of scanning traffic generated by the numerous subprograms running in parallel, a large
amount of broadcast request traffic is generated by each infected system. If several machines
on a local network segment are infected, then the resulting attempt to propagate the infection
to their neighbors simultaneously can generate broadcast requests at “flooding” rates.
Systems on the receiving end of an effective “broadcast flood” may experience the effects of a
DoS attack.
Code Red II allows remote attackers and intruders to execute arbitrary commands on infected
Windows 2000 systems. Compromised systems are then subject to files being altered or
destroyed. This creates a DoS condition for entities that may be relying on the altered or
destroyed files. Furthermore, compromised systems are also at high risk for being exploited to
generate other types of attacks against other servers.
What do you do if you’re infected?
The propagation function, the widespread, automated attack and propagation characteristics of
the Code Red II may cause bandwidth DoS conditions in isolated portions of the network,
particularly near groups of compromised hosts where Code Red II is running.
Several antivirus software vendors have created tools that remove the harmful affects of the
worm and restore the computer to a seemingly normal working configuration. This fix, however, is
useless if the infected computer had been accessed by an attacker who installed other backdoors
on the system that would be unaffected by the Code Red II patch tool.
According to FedCIRC (Federal Computer Incident Response Capability), due to the malicious
actions of this worm, patching and rebooting an infected server will not solve the problem. The
system’s hard drive should be reformatted and all software should be reinstalled.
Technical Details of the Code Red II Worm
The Code Red II worm also has three phases – preparation, propagation, and Trojan insertion. Based upon current analysis, Code
Red II only affects Web servers running on the Microsoft Windows 2000 operating system platform.
Phase 1: Preparation
During the preparation phase, the worm checks the current date to determine whether it will run
at all. If the date is later than October 1, 2002, then the worm will cease to function and will
remain infinitely dormant. If the date is before October 1, 2002, then all functions will be
performed. Although this discovery may bring hope that after October 1, 2002, this worm will no
Page 13
GAO-01-1073T
Question
Answer
longer be a threat, this date constraint can be easily changed in a variant. The other activities
conducted during the preparation phase include:
The functionality of Code Red II is dependent on both the system’s environment and the
current date. Code Red II checks the default system’s language, e.g., English, Chinese, etc.,
and stores that information.
The worm also checks if the system has been previously infected, by searching for the
existence of a specific file. If the file exists, then Code Red II becomes dormant and does not
re-infect the system. If the file does not exist, Code Red II creates the file and continues the
process.
Preparation is finalized when the worm disables the capability of the Windows 2000 operating
system to repair itself if it discovers that one of its required system files has been modified in
any way. This becomes important during the Trojan Insertion function.
Phase 2: Propagation
Once the worm has completed the preparation phase, it immediately starts the propagation and
Trojan insertion phases to complete infection.
Code Red II creates hundreds of subprograms to propagate itself. The number of subprograms
created depends upon the default language that the worm identified in the Preparation function. If
the system’s default language is Chinese, then 600 subprograms are created. If the default
language is not Chinese, then 300 subprograms are generated.
The propagation phase is unique because Code Red II seeks to copy itself to computers that are
near the infected system. The algorithm uses the infected system’s own Internet address to
generate a list of random Internet addresses. The generated list is comprised of Internet
addresses that are closely related to the infected system. The rationale is that similar systems
should reside in the “neighborhood” of the infected system, resulting in an increased chance of
infection.
Phase 3: Trojan Insertion
Each of the subprograms is tasked with scanning one of the randomly generated Internet
addresses to identify and infect another vulnerable system. Like Code Red, this worm uses the
buffer overflow vulnerability to infect its target. Once a new target is infected, the process
continues.
Code Red II is more malicious than the Code Red worm discussed earlier, due to the existence
of the Trojan horse backdoor programs that Code Red II leaves behind on the infected computer.
The basic process follows:
Initially, executable files are copied to specific locations on the Web server, which by
necessity, are accessible by any remote user. These executable files can run commands sent
by a remote attacker to the server through the use of well-crafted Web commands.
A Trojan horse program is planted on the server that allows further exploit of the infected
computer. The Trojan horse program is named after a required system program that executes
each time the system is rebooted. It is also placed in a location that ensures that the Trojan
horse program will be run instead of the required system program at reboot. Upon execution,
the Trojan horse changes certain system settings that grant remote attackers read, write, and
execute privileges on the Web server.
Twenty-four hours after the preparation function is initiated, Code Red II forces the infected
system to reboot itself. Although the reboot eliminates the memory resident worm, the
backdoor and the Trojan horse programs are left in place since they are stored on the
system’s disks. The reboot also restarts the IIS software, which, in turn, ensures that the Web
server uses the newly compromised system settings.
Since the Trojan horse will always be executed each time the system reboots, Code Red II
guarantees that remote attackers will always have access to the infected system. This is
Page 14
GAO-01-1073T
Question
Answer
important, since even if the executable files copied at the beginning of the Trojan Insertion phase
are deleted, the excessive privileges the Trojan sets at reboot are still in place. Therefore, the
Trojan enables a remote attacker to perform similar exploits using these excessive privileges.
SirCam
Question
What is it?
Answer
SirCam is a malicious computer virus that spreads through E-mail and potentially through
unprotected network connections. What makes SirCam stealthy is that it does not rely on the Email capabilities of the infected system to replicate. Other viruses, such as Melissa and
ILOVEYOU, used the host machine’s E-mail program while SirCam contains its own mailing
capability.
Once the malicious code has been executed on a system, it may reveal or delete sensitive
information.
How does it spread?
SirCam was first detected on July 17.
This mass-mailing virus attempts to send itself to E-mail addresses found in the Windows
Address Book and addresses found in cached files.
It may be received in an E-mail message saying “Hi! How are you?” and requesting help with an
attached file. The same message could be received in Spanish.
Since the file is sent from a computer whose owner is familiar enough with the recipient to have
their E-mail address in their address book, there is high probability that the recipient will trust the
attachment as coming from a known sender. This helps ensure the virus’s success in the wild and
is similar to the social engineering approach used by Melissa and I Love You.
The E-mail message will contain an attachment that will launch the code when opened. When
installed on a victim machine, SirCam installs a copy of itself in two hidden files. It then “steals”
one of the target system’s files and attempts to mail that file with itself as a Trojan, that is, a file
with desirable features, to every recipient in the affected system’s address book. It can also get
E-mail addresses from the Web browser.
Who is at risk?
What damage can it do?
SirCam can also spread to other computers on the same Windows network without the use of Email. If the infected computer has read/write access to Windows network computers, SirCam
copies itself to those computers, inevitably infecting the other computer.
Any E-mail user or any user of a PC with unprotected network connections that is on the same
Windows network as an infected computer.
SirCam can publicly release sensitive information and delete files and folders. It can also
completely fill the hard drive of the infected computer. Furthermore, it can also lead to a decrease
in the speed of the Internet.
Specifically,

It can cause security breaches by attaching randomly chosen documents to itself and then Emailing them to other parties. This allows the worm to cause unauthorized disclosure of
confidential information.
Page 15
GAO-01-1073T
Question
Answer

It can also delete files and folders. There is a one in fifty chance that an infected computer
will either have its hard drive erased or completely filled with garbage on October 16.

It can create a file named C:\Recycled\sircam.sys which consumes all free space on the C:
drive. A full hard drive prevents users from saving files to that drive, and in certain
configurations impedes system-level tasks, such as auditing and printing.
It can result in a denial of service attack by flooding E-mail systems with useless E-mail
containing attachments of various sizes.
What do you do if you’re
Most anti-virus software vendors have released updated information, tools or virus databases to
infected?
help detect and partially recover from SirCam.
Technical Details of the SirCam
Virus
SirCam detaches itself from the E-mail attachment and attempts to execute its program file on
Actions performed once the user
the target machine.
executes the attachment
It copies itself to several directories on the target system.
It then “steals” one of the target system’s files and attempts to mail that file with itself as a
Trojan to every recipient in the affected system’s address book. It can also get E-mail
addresses from the Web browser. The subject line and the attachment’s name differ from email to e-mail. The attached file is where the virus’ malice lies: the infected E-mail’s attachment
has a name that matches the subject line and two extensions, the second being .exe or .bat or
.com. For example, a Word file called SAMPLE.DOC could be attached to the E-mail as
SAMPLE.DOC.EXE.
It can also delete files and folders. There’s a one in fifty chance that an infected computer will
either have its hard drive erased or completely filled with garbage on October 16.
In addition to E-mail propagation, SirCam can copy itself to other computers on the Windows
network that have write-able access. SirCam will copy itself to those computers and rename itself
to be a system file that will be executed upon the next system reboot.
a
Buffer overflows occur when programs do not adequately check input for appropriate length. Thus, any unexpected
input “overflows” onto another portion of the central processing unit’s executions stack. If this input is chosen judiciously
by a rogue programmer, it can be used to launch code of the programmer’s choice.
b
http://www.cert.org/advisories/CA-2001-19.html
(460513)
Page 16
GAO-01-1073T